Beylik period in western Anatolia: A case study from Ephesus and Miletus

JASREP-00255; No of Pages 11 Journal of Archaeological Science: Reports xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Journal of Archaeological Science: Reports journal homepage: www.elsevier.com/locate/jasrep

Moulded wares production in the Early Turkish/Beylik period in Western Anatolia: A case study from Ephesus and Miletus Sylvie Yona Waksman a,⁎, Jacques Burlot a, Beate Böhlendorf-Arslan b, Joanita Vroom c a b c

CNRS, UMR 5138 "Archéométrie et Archéologie", Maison de l'Orient et de la Méditerranée, 7 rue Raulin, 69365 Lyon cedex 7, France Römisch-Germanisches Zentralmuseum, Leibniz-Forschungsinstitut für Archäologie, Ernst Ludwig Platz 2, 55116 Mainz, Germany Faculty of Archaeology, Leiden University, P.O. Box 9514, 2300 RA Leiden, The Netherlands

a r t i c l e

i n f o

Article history: Received 7 May 2015 Received in revised form 14 July 2015 Accepted 12 November 2015 Available online xxxx Keywords: Turkish period Western Turkey Pottery Moulded wares Chemical analysis Provenance Technology

a b s t r a c t The beginning of the Turkish period in Western Anatolia initiates new ceramic types, related to the “Islamic” world, in the local pottery repertoire. To study this evolution from Byzantine to Turkish times, our paper focuses on one of these types, moulded wares, and on two sites, Miletus and Ephesus. At both sites, evidence of local production was found, which enabled us to define or re-define the corresponding chemical reference groups using WD-XRF. The identification of the origin of the finds, or their attribution to as yet unlocated workshops whose diffusion may be approached, provide solid grounds for discussing the introduction of moulded wares in Western Anatolia. SEM analyses give some preliminary insight into the emergence and development of new ceramic techniques in the region, whose “end-products” are represented by the well-known Iznik Ware. This paper constitutes a first step in research carried out in the framework of the POMEDOR project (People, Pottery and Food in the Medieval Eastern Mediterranean). © 2015 Published by Elsevier Ltd.

1. Introduction Although long neglected, the Byzantine and Turkish archaeological contexts which precede the Ottoman period in Western Anatolia gradually tend to be both better documented, and better studied for their pottery (Spieser, 1996; Böhlendorf-Arslan, 2004; Vroom, 2005; Mania, 2006; Pfeiffer-Taş, 2010, 2011; Gök Gürhan, 2011; Doğer, 2013; Mercangöz, 2013; Ladstätter, 2015; Vroom, 2015; Vroom and Fındık, 2015). In parallel, the development of archaeometric research has provided tools for more comprehensive approaches to production and diffusion of Western Anatolian ceramics in these periods (Scott and Kamilli, 1981; Waksman, 1995; Demirci et al., 1996; Okyar, 2010; Okyar et al., 2011; Kırmızı, 2012; Budak Ünaler, 2013; Özçatal et al., 2014), especially in recent years thanks to the specialized chemical database developed in Lyon (Waksman and François, 2004–2005; Sauer and Waksman, 2005; Waksman and Wartburg von, 2006; Waksman, 2015). This paper focuses on a specific type of ceramics, moulded wares, which appear in Western Anatolia during the early Turkish period, and on two sites, Ephesus and Miletus (Fig. 1), where moulded wares found together with moulds provide evidence for local production

⁎ Corresponding author. E-mail address: [email protected] (S.Y. Waksman).

(Vroom, 2005; Böhlendorf-Arslan, 2008). Although moulded wares were already manufactured during the Hellenistic and Roman periods (Hochuli-Gysel, 1977), they appear in Western Turkey again much later, together with other types of pottery, such as turquoise glazed wares, which can be related to products from the Islamic world. They are dated there to the 14th c. AD, which closely follows the arrival of Turkish rulers and populations. The latter may have brought their pottery traditions and techniques, together with their dining habits and diet, taste and symbolic uses (Böhlendorf-Arslan, 2002; Dalaman, 2015; Vroom, 2015). Such issues are investigated in the larger framework of the POMEDOR project.1 In this paper, we focus on the introduction of moulded wares in the local pottery repertoire in Ephesus and Miletus, raising issues of provenance and technology, and attempting to contextualize this introduction. Provenances were investigated by chemical analysis using WDXRF (CNRS UMR 5138), building on previous archaeometric research on Ephesus (Sauer and Waksman, 2005; Waksman, 2014, 2015) and taking the opportunity to update previous views. Technical traditions were studied under the SEM (CTμ, Lyon), with special interest in paste/slip/ glaze associations.

1 POMEDOR project “Population, Pottery and Food in the Medieval Eastern Mediterranean”, http://www.pomedor.mom.fr.

http://dx.doi.org/10.1016/j.jasrep.2015.11.015 2352-409X/© 2015 Published by Elsevier Ltd.

Please cite this article as: Waksman, S.Y., et al., Moulded wares production in the Early Turkish/Beylik period in Western Anatolia: A case study from Ephesus and Miletus, Journal of Archaeological Science: Reports (2015), http://dx.doi.org/10.1016/j.jasrep.2015.11.015

2

S.Y. Waksman et al. / Journal of Archaeological Science: Reports xxx (2015) xxx–xxx

times. During that period the site seemed to have been used as some sort of residence or living quarters. Numerous fragments of tableware, glass vessels and metal objects were found in large quantities, often of high quality unrivalled for that era. Among the finds are fragments of blue-green Celadon bowls and plates from China, Islamic fritware with vegetal motives in blue paint, moulded jugs covered with a micaceous coating or a glaze, monochrome and polychrome Sgraffito Wares, unglazed amphorae and cooking pots. Most of the moulded fragments sampled for this article originate from Sondage 2, a large excavation trench situated along the western part of the monument. They can be generally dated in the 13th–14th c. and sometimes more specifically in the 2nd half of the 14th c. The documentation of the material from the Türbe and the Isa Bey Hamam is nearly finished and awaits publication, while the study of the Tribune material is still ongoing. Besides basic analyses of the pottery, glass and small finds from the three sites to date the stratigraphy, the project aims to analyse and revalue the current knowledge of ceramic assemblages. This will include spatial analyses of the finds within the rooms (architectural units) of the sites displaying the function of the separate spaces through time. Finally, the study so far has shown new pottery types, new regional production sites and new patterns of trading networks (Vroom, 2015).

Fig. 1. Location of the sites mentioned in the text.

2. Archaeological sites, contexts and sampling 2.1. Sites and contexts 2.1.1. Recent ceramic research on Medieval Ephesus Ephesus is continuously inhabited from Hellenistic to Ottoman times. In its last occupation phases the centre of habitation extended from the ancient city (situated near the harbour) towards the Ayasoluk Hill. In fact, the Beylik and Ottoman influence on the city and the conquest of its lands in 1304 initiated a new period and added landmarks to the urban landscape that last to today. The excavation of Ephesus by the Austrian Archaeological Institute has been ongoing since 1893, but interest in the late Byzantine to Ottoman periods has only lately been shown at the initiative of Ladstätter, the current director of the excavations. The current study, executed by Vroom and her team from Leiden University, focuses on finds from four sites in particular: the Artemis Temple (also known as “Artemision”), the Isa Bey Hamam, an anonymous Türbe and the socalled Tribune (e.g. Vroom, 2015). Recent excavations produced a considerable amount of material from these last three sites, which makes a longue durée research focus possible. The excavation of an anonymous Türbe, a funerary monument south to the Artemis Temple, took place from 2009 to 2012 (Ladstätter, 2015). This Türbe dates from the early 15th century, and was erected cutting through several previous structures: a graveyard, a Byzantine wall of two phases and a pottery kiln (Parrer, 2015). The ceramic finds mostly range in date from Middle Byzantine to more recent times (circa 10th–early 20th centuries), with a few earlier exceptions. The typochronology as well as the provenance of the Türbe ceramic finds have been discussed, based on a simple functional ordering in glazed wares, unglazed wares and some potter's tools (Vroom and Fındık, 2015; Waksman, 2015). The results eventually shed light on the production processes and trade relations of Medieval and Ottoman Ephesus (or “Ayasoluk”) in general, and on the dating of the anonymous monument in particular. Some hundred metres westwards from this Türbe lays the Tribune site. Here excavation from 2009 to 2011 revealed an additional motivation of continuous habitation based on the abundance and wealth of the Tribune's material record dating from Byzantine to Beylık and Ottoman

2.1.2. Early Turkish Miletus After the conquest of Miletus by the Turkish dynasty of Menteşeoǧluları in 1261, life in the city continued without any interruption. Thereafter an increased construction activity began. The castle was renovated; now it served to secure the two ports that were also used for long distance trade with Venice. In that period a Venetian consulate, trading houses and a new church was constructed (Müller-Wiener, 1967, 288-90; Zachariadou, 1983). In the 14th century a pottery workshop was operating in the remains of the so-called Bishop’ Palace north of the Michael's church (Müller-Wiener et al., 1979, 170). More pottery workshops were probably located on the road west of the ancient bouleuterion (Sarre, 1935, 70; Wulzinger et al., 1935, 57). With the construction of visitor paths that lead from the Byzantine theatre castle hill to the Michael's Church and the ancient city centre, several older excavation rubble mounds were removed. The rubble may have been created during excavation of the workshop area. From these many tripods, misfired pottery and other ceramic finds originated (Böhlendorf-Arslan, 2008, 375: “STR”, “WO” or “W”). Other ceramic finds are from surface cleaning of the theatre castle (Böhlendorf-Arslan, 2008: “TK” or “TO”) or found during the renovation work in the theatre (Böhlendorf-Arslan, 2008: “T”). The excavations at the temple of Athena in the south of the city and the Michael's Church also provide Byzantine and early Ottoman material in the upper layers, which are not fully published yet (BöhlendorfArslan, 2008: “AT” or “MK”). Although these finds do not come from stratified contexts, they were of particular interest for our study as they include evidence of pottery production associated to pottery of the early Turkish period.

2.2. Sampling The sampling considered here2 includes three categories of ceramics and clay objects, selected among a larger repertoire studied in the framework of the POMEDOR project: 2 Samples from Miletus in Böhlendorf-Arslan, 2008: BZY979 (Kat. 6, Abb. 4); BZY980 (Kat. 54, Abb. 14.54), BZY984 (Kat. 7, Abb. 5.7), BZY985 (Kat. 10, Abb. 5.10), BZY986 (Kat. 8, Abb. 5.8), BZY989 (Kat. 9, Abb. 5.9), BZY995 (Kat. 14, Abb. 6.14). Samples from Ephesus in Sauer and Waksman, 2005: BYZ439 (EPH-ART 17), BYZ440 (EPH-ART 18), BYZ441 (EPH-ART 19), BYZ442 (EPH-ART 20); in Waksman, 2015: BZY332, BZY333 (Fig. 2), BZY373 (Figs. 1 and 3); in Vroom and Fındık, 2015: BZY332 (cat. 212), BZY373 (cat. 144); in Vroom, 2005: BZY966 (No. 32, Pl. 4, 11).

Please cite this article as: Waksman, S.Y., et al., Moulded wares production in the Early Turkish/Beylik period in Western Anatolia: A case study from Ephesus and Miletus, Journal of Archaeological Science: Reports (2015), http://dx.doi.org/10.1016/j.jasrep.2015.11.015

S.Y. Waksman et al. / Journal of Archaeological Science: Reports xxx (2015) xxx–xxx

– samples taken as references for local production: kiln furniture, especially tripod stilts used as spacers between glazed wares in pottery kilns (samples BZY971-976 from Miletus; from Ephesus: BZY964-965, 967 [Artemision], BZY332-333 [Türbe], BZY901-902, 958 [Tribune]); biscuit-fired wares, i.e. unfinished wares discarded between the first and second firings, before the application of the glaze (BZN 55-56, BZY980-982 from Miletus; BZY903, 905 from Ephesus [Tribune]); overfired wasters (BZY979 from Miletus, stuck to a tripod stilt) (Fig. 2); – moulded wares: the latter are usually jugs covered with a shiny, micaceous coating (BZY991-996 from Miletus; from Ephesus BYZ439-442 [Artemision], BZY935-943 [Tribune]). A few examples are glazed (BZY997 from Miletus; from Ephesus BZY373 [Türbe], BZY931-934, 956 [Tribune]) (Figs. 3–4); – moulds (BZY984-990 from Miletus; from Ephesus BZY950 [Tribune], BZY966 [Artemision]), some of which bear the same motives as the moulded wares considered (Figs. 3–4). – All samples were analysed by WD-XRF for provenance issues, a subsampling corresponding to the glazed moulded wares were further investigated for their technology through SEM analysis of slips and glazes.

3. Materials and methods 3.1. Chemical analysis of ceramic bodies WD-XRF analyses were carried out in Lyon (CNRS UMR 5138) using a Bruker SRS 3400 spectrometer, then a Bruker S8 Tiger from 2014 onward. Samples are cut out with a diamond-coated saw, in a way which alters neither the form nor the decoration of the sherd. Glaze and slip, when present, and an external layer, whose chemical composition is more liable to be altered by the burial environment, are removed. After heating at 950 °C (removal of water, volatile elements, organics), cooling and grinding, 800 mg of ceramic powder is mixed with 3200 mg of flux (lithium metaborate and tetraborate). The mix is heated to liquid state in a gold and platinum crucible, then cast into a bead. Analyses are carried out on these homogeneous beads, of fixed

3

geometry, which correspond to a mean chemical composition representative of the initial material. Twenty-four elements are quantified, after calibration of the set-up using 40 geological standards (CRPG, USGS, NIST, British Chemical Standards, etc.). The calibration is regularly checked using 3 in-house pottery standards. 3.2. Cluster analysis Out of the 24 elements quantified, 17 are taken as active variables in multivariate statistical treatments used to classify samples into groups of similar chemical composition. These include major and minor elements in ceramics (MgO, Al2O3, SiO2, K2O, CaO, TiO2, MnO, Fe2O3) and trace elements having various geochemical behaviour (V, Cr, Ni, Zn, Rb, Sr, Zr, Ba, Ce). Classifications are carried out by hierarchical cluster analysis, applied to standardized data, using euclidian distance and average linkage (e.g. Picon, 1984). Interpretation however requires further examination of the initial individual chemical compositions, taking into account various geochemical, technological and analytical factors (Waksman, forthcoming). 3.3. SEM analysis Samples were embedded in a polyester resin, cut in a cross section in order to expose all the layers (ceramic body, slip and glaze), polished with a 0.25 μm diamond paste and then coated with a layer of carbon. Prior to coating with carbon, the cross-sections were observed under a binocular microscope with magnifications in the range 10× to 115×. Slips and glazes analyses were carried out with a SEM (FEI Quanta FEG 250) at the “Centre Technologique des Microstructures” (CTμ) in Lyon. The study of the microstructures was mainly performed using backscattered electron images (BSE images) which allow differentiating the various phases according to their atomic number. Elemental compositions were determined by energy-dispersive X-ray spectrometry (EDS) using a silicon drift detector 30 mm2 SAMx. All measurements were operated at 15 kV during 60 s. Standardless quantification was performed using a PAP correction method of the intensities. IDFix software was used for data acquisition and evaluation, and the reliability of the results was tested by measuring glass and geological reference standards (Corning Brill B, C and D, DR-N). Concentrations are given

Fig. 2. Examples of samples taken as local references in Miletus and Ephesus: tripod stilts (BZY901-902, from Ephesus); overfired waster stuck to a tripod stilt (BZY979, from Miletus); unfinished, biscuit-fired wares (BZY980, 982, BZN 55–56, from Miletus).

Please cite this article as: Waksman, S.Y., et al., Moulded wares production in the Early Turkish/Beylik period in Western Anatolia: A case study from Ephesus and Miletus, Journal of Archaeological Science: Reports (2015), http://dx.doi.org/10.1016/j.jasrep.2015.11.015

4

S.Y. Waksman et al. / Journal of Archaeological Science: Reports xxx (2015) xxx–xxx

Fig. 3. Moulded wares and moulds belonging to chemical group “Miletus local”.

as average of three to five measurements on different zones selected within the regions of interest, avoiding the altered ones and the slip or body interfaces. 4. Results and discussion 4.1. Provenance issues The identification of Ephesus and Miletus local products was based on the classification of the samples according to the chemical composition of their body. It distinguishes three main chemical groups, some samples being either unclassified or marginal to one of the groups (Fig. 5, Table 1). A first group corresponds to moderately calcareous pastes showing strong ultrabasic features (high magnesium, chromium, nickel) (Table 1). It has fairly variable concentrations in these elements, as is often observed in such geochemical contexts (Waksman, forthcoming), as well as in several others such as potassium and iron. But local references samples within the group (Fig. 5, red circles) clearly identify this group as local to Miletus, and indicate that the range of chemical concentrations observed is within the intra-group variations. This group only gathers samples found in Miletus, including all the moulds (Fig. 3). However, most of the moulded wares found in Miletus are part of another group (Fig. 4, middle), which includes samples from both Ephesus and Miletus and corresponds to Ephesus group b/2 (or b, or 2) as defined previously (Sauer and Waksman, 2005; Waksman, 2015). Although it also has, to a lesser extent, some ultrabasic features, this group clearly differs chemically from the previous one (lower calcium, chromium, higher potassium and aluminium contents, etc., Table 1). The third main group corresponds to our previous Ephesus group c/4, also a group of low calcareous pastes distinguished by their high iron and fairly high aluminium contents (Table 1). Local references from Ephesus (Fig. 5, red triangles) are present in group c/4 only, which also includes the two moulds found there (Fig. 4, top). Ephesus group b/2 includes neither local reference samples stricto sensu, nor moulds, and gathers samples from the two sites. During earlier stages of research, the association of moulded wares and moulds in the contexts of reoccupation of the Artemision in Ephesus had been considered as strong enough evidence for their local

production (Vroom, 2005; Sauer and Waksman, 2005) and had lead us to consider both Ephesus groups b/2 and c/4 as local. The present study gives the opportunity to revise these early views, taking into account the fact that both moulds and moulded wares may have circulated, a feature well known for the Hellenistic and Roman periods (e.g. Picon and Lasfargues, 1974; Rotroff and Oliver, 2004) but still little acknowledged for the Medieval one. The present research questions the local status of Ephesus group b/2, although petrographic and heavy minerals analyses, carried out by Sauer, suggest that it originates from Ephesus or its surroundings (Sauer and Waksman, 2005, 54). It may thus correspond to another production of the region, although not necessarily located in Ephesus. A few other samples not included in these three main groups (Fig. 4, bottom) may be related to wares previously studied in Ephesus. It is the case of moulded wares showing a dark red fabric, more typical of cooking wares (BZY935-937). Two of them (BZY935-936) are close chemically to late Roman cooking wares and Medieval common wares, considered local to Ephesus as they constitute one of the main fabric there at various periods3 (Sauer and Waksman, 2005: group D; Waksman, 2014). Another marginal sample, a piece of dark green glazed moulded ware (BZY933), is also close to a previously analysed tripod stilt from Ephesus (Sauer and Waksman, 2005: BYZ449, not shown in Fig. 5), except for its higher contents in ultrabasic components (Mg, Cr, Ni). Other glazed examples are spread among the three main groups (3 related to Ephesus b/2 group, 2 to Ephesus c/4 and 1 in Miletus group). Archaeometric research on Western Anatolian sites such as Ephesus and Miletus had previously mainly focused on the Greco-Roman periods (e.g. Jones, 1986; Hughes et al., 1998; Zabehlicky-Scheffenegger et al., 1996; Schneider, 2000; Akurgal et al., 2002; Schneider and Japp, 2009; Mommsen and Japp, 2009). The later periods have the advantage of abundant and – in most cases – unambiguous local reference material, which gives the opportunity to define chemical reference groups. The diachronic use of reference groups is however not straightforward, as we already pointed out in the case of Ephesus and Pergamum, partly due to the geological and geochemical variability of Western Anatolia (Waksman, 2005, 2014). Clay procurement may be multiple at a given

3

We would like to thank L. Peloschek (OAI, Vienna) for confirming this point.

Please cite this article as: Waksman, S.Y., et al., Moulded wares production in the Early Turkish/Beylik period in Western Anatolia: A case study from Ephesus and Miletus, Journal of Archaeological Science: Reports (2015), http://dx.doi.org/10.1016/j.jasrep.2015.11.015

S.Y. Waksman et al. / Journal of Archaeological Science: Reports xxx (2015) xxx–xxx

5

Fig. 4. Moulded wares and moulds belonging to chemical groups “Ephesos c/4” (local) (top); “Ephesus b/2” (regional?) (middle); other samples (down).

Fig. 5. Classification according to chemical analysis of samples from Miletus and Ephesus. Samples are identified by their laboratory number, symbols indicate find spots and types, and point out reference samples for local production. The main chemical groups are underlined.

Please cite this article as: Waksman, S.Y., et al., Moulded wares production in the Early Turkish/Beylik period in Western Anatolia: A case study from Ephesus and Miletus, Journal of Archaeological Science: Reports (2015), http://dx.doi.org/10.1016/j.jasrep.2015.11.015

6

S.Y. Waksman et al. / Journal of Archaeological Science: Reports xxx (2015) xxx–xxx

Table 1 Chemical compositions of samples from Miletus and Ephesus, ranked as in the dendrogram shown in Fig. 5, together with comparative data. Lab nb.

Desc.

CaO

Fe2O3

TiO2

K2O

SiO2

Al2O3

MgO

MnO

(Na2O)

(P2O5)

Zr

Sr

7.19 7.22 6.80 7.12 7.19 7.20 7.30 6.86 6.99 6.96 7.41 6.57 6.74 7.41 7.30 7.55 7.07 7.90 6.29 7.93 7.33 8.45 8.02 7.51 7.26 0.48

0.682 0.687 0.681 0.695 0.681 0.667 0.693 0.711 0.705 0.673 0.679 0.673 0.714 0.739 0.688 0.707 0.673 0.729 0.642 0.785 0.746 0.680 0.695 0.678 0.696 0.030

2.84 2.71 2.13 2.13 1.94 1.87 1.92 2.65 2.63 2.94 2.96 2.99 3.12 3.07 2.97 2.97 3.03 3.38 3.00 3.01 2.92 2.56 2.22 2.83 2.70 0.43

54.15 53.80 54.98 54.45 52.93 52.17 53.04 56.93 56.01 57.01 55.86 58.52 57.23 56.55 53.53 54.37 55.44 54.46 58.85 55.51 54.82 49.84 50.36 52.42 54.72 2.27

14.82 14.77 14.38 15.06 14.84 14.84 15.03 14.58 14.64 14.68 15.16 14.45 14.40 15.46 15.10 15.24 15.07 16.13 14.95 16.26 15.61 16.00 15.55 15.10 15.09 0.52

6.73 6.93 6.17 6.65 6.78 6.90 7.02 5.90 6.30 5.98 6.04 5.12 5.09 5.46 5.92 6.06 5.64 5.53 4.93 4.38 4.98 7.62 7.47 6.75 6.10 0.85

0.1022 0.1051 0.0979 0.1001 0.0940 0.1119 0.0952 0.0920 0.0987 0.0949 0.1031 0.0930 0.0866 0.1009 0.1057 0.1104 0.0991 0.1027 0.0898 0.1018 0.1031 0.1157 0.1170 0.1394 0.1025 0.0110

1.83 1.94 2.00 2.27 2.08 2.01 2.71 1.82 1.78 1.52 1.54 1.54 1.63 1.47 1.33 1.35 1.35 1.44 1.97 1.05 1.36 1.57 1.91 1.48 1.71 0.37

0.24 0.25 0.24 0.27 0.22 0.37 0.30 0.27 0.28 0.24 0.28 0.43 0.35 0.35 0.49 0.43 0.26 0.37 0.40 0.64 0.74 0.26 0.25 0.31 0.34 0.13

196 196 183 182 181 172 183 199 198 175 173 210 198 204 180 189 188 193 192 240 220 150 159 176 189 19

273 268 276 243 315 324 283 248 253 264 272 288 294 284 351 300 318 293 267 286 311 281 316 315 288 26

Similar to Ephesus local group c/4 BZY901 lrt 1.27 12.16 BZY950 mo 1.47 11.83 BZY966 mo 1.34 12.35 BZY903 lrb 1.32 12.20 BZY905 lrb 1.30 11.84 BZY902 lrt 1.21 11.24 BZY934 gmw 1.21 11.45 BZY932 gmw 1.30 11.30 BZY964 lrt 1.20 11.22 BZY333 lrt 1.54 11.17 BZY334 lro 1.66 10.81 BZY332 lrt 1.75 10.88 BZY958 lrt 1.76 10.64 BZY943 mw 1.50 11.54 BZY965 lrt 1.25 11.03 m 1.41 11.44 σ 0.20 0.53

1.043 1.015 1.033 1.031 1.044 1.038 1.040 1.051 1.024 1.007 0.980 0.969 1.001 0.974 1.022 1.018 0.027

4.36 4.39 4.31 4.15 4.22 4.35 4.25 4.30 4.35 4.00 4.13 4.86 4.51 3.95 3.85 4.27 0.24

51.26 51.26 50.95 51.42 51.99 52.92 52.82 52.92 52.38 53.30 53.19 53.36 52.06 53.45 54.53 52.52 1.02

24.22 23.89 24.53 24.17 23.98 23.66 23.61 23.68 23.73 23.11 23.00 22.31 23.30 22.25 22.71 23.48 0.69

3.57 3.62 3.51 3.57 3.43 3.49 3.62 3.33 3.58 4.08 3.98 4.02 3.53 4.38 3.45 3.68 0.30

0.1404 0.1460 0.1502 0.1491 0.1448 0.1320 0.1355 0.1253 0.1352 0.1517 0.1306 0.1426 0.1370 0.0967 0.1023 0.1346 0.0162

1.38 1.81 1.30 1.48 1.35 1.39 1.31 1.33 1.40 1.06 1.52 1.21 1.53 1.35 1.46 1.39 0.17

0.27 0.28 0.29 0.26 0.30 0.26 0.27 0.26 0.26 0.29 0.29 0.26 0.31 0.27 0.31 0.28 0.02

226 231 229 223 231 247 241 257 231 214 203 200 221 176 209 223 20

Marginals to Ephesus local group c/4 BZY991 mw 3.87 10.59 0.970 BZY933 gmw 2.60 10.13 1.027

3.96 3.51

52.58 55.32

21.87 21.42

4.25 3.91

0.1220 0.1016

1.29 1.47

0.24 0.24

Similar to Ephesus group b/2 BZY995 mw 4.88 BZY993 mw 4.59 BZY940 mw 4.06 BZY942 mw 4.48 BYZ441 mw 4.95 BZY938 mw 4.93 BZY994 mw 3.61 BZY956 gmw 3.03 BZY931 gmw 3.44 BZY373 gmw 3.14 BYZ442 mw 4.01 BZY941 mw 4.74 BZY939 mw 3.42 BYZ440 mw 3.32 BZY996 mw 3.23 m 3.99 σ 0.72

8.68 8.56 8.62 8.79 8.33 9.43 8.30 8.36 8.81 8.16 8.33 8.10 8.63 8.29 8.25 8.51 0.34

0.938 0.940 0.977 0.943 0.915 0.962 0.963 0.976 0.993 0.981 0.968 0.921 0.996 1.024 0.998 0.966 0.031

3.97 3.92 4.15 3.97 3.88 3.82 3.97 3.92 3.87 3.86 3.59 3.79 4.05 4.21 4.03 3.93 0.15

54.66 55.60 55.58 55.98 55.57 53.81 56.98 57.99 56.52 57.81 57.70 57.15 56.84 55.53 56.90 56.31 1.20

20.56 20.36 20.80 20.39 20.25 20.70 20.44 20.35 20.69 20.57 19.69 19.53 21.00 21.98 20.98 20.55 0.57

4.38 4.14 4.06 3.64 3.84 4.41 3.57 3.52 3.94 3.52 3.93 3.86 3.02 3.58 3.54 3.80 0.37

0.1006 0.1008 0.1014 0.1037 0.1043 0.1110 0.0989 0.0975 0.1027 0.0911 0.0928 0.0974 0.1089 0.0929 0.0941 0.0999 0.0058

1.35 1.31 1.18 1.22 1.45 0.97 1.54 1.29 1.15 1.35 1.22 1.28 1.40 1.49 1.51 1.31 0.15

Unclassified BYZ439 mw BZY967 lrt BZY935 mw BZY936 mw BZY937 mw

7.47 6.30 9.35 9.91 9.26

0.980 0.673 1.336 1.304 0.943

3.23 4.66 2.48 3.16 2.03

61.19 51.16 62.48 60.20 63.24

17.56 21.99 20.09 20.50 18.88

2.89 2.54 1.87 1.85 2.67

0.0894 0.0647 0.0811 0.1141 0.0839

1.40 0.67 0.98 0.81 1.72

Miletus, local group BZY984 mo 11.17 BZY986 mo 11.32 BZY973 lrt 12.15 BZY975 lrt 10.59 BZY985 mo 13.00 BZY971 lrt 13.53 BZY989 mo 11.40 BZY976 lrt 9.94 BZY979 lro 10.30 BZY974 lrt 9.57 BZY981 lrb 9.73 BZN 56 lrb 9.35 BZY980 lrb 10.37 BZN 55 lrb 9.13 BZY997 gmw 12.23 BZY983 mo 10.85 BZY992 mw 11.12 BZY988 mo 9.69 BZY977 lro 8.61 BZY982 lrb 10.05 BZY987 mo 11.13 BZY972 lrt 12.58 BZY978 lro 13.12 BZY990 mo 12.54 m 10.98 σ 1.37

4.70 11.26 0.95 1.49 0.90

Rb

Zn

Cr

Ni

(La)

Ba

V

Ce

118 115 109 *96 118 113 *60 120 121 114 129 121 117 128 120 120 124 131 114 131 131 125 126 121 121 6

81 86 78 89 84 88 85 88 85 87 91 93 91 98 85 97 89 100 80 113 103 99 96 97 91 8

392 382 378 398 361 359 374 408 403 371 351 351 395 388 377 382 333 368 336 395 389 369 371 357 375 20

440 443 403 438 427 447 429 388 405 404 418 351 358 396 417 447 400 425 323 353 372 543 494 463 416 48

36 52 39 31 35 45 35 34 39 36 36 37 40 42 41 41 44 49 49 46 47 57 41 48 42 6

480 454 464 495 476 465 468 511 490 565 524 538 576 586 615 615 606 718 681 580 525 447 523 512 538 72

101 103 105 111 103 107 117 102 103 105 105 100 108 108 121 116 106 113 95 122 109 114 116 111 108 7

83 88 82 89 83 78 86 82 85 73 79 82 79 81 87 86 92 87 80 101 92 95 94 89 86 6

160 163 165 156 146 164 162 164 143 155 161 164 119 136 165 155 13

168 164 172 169 155 168 167 160 142 162 160 164 *100 148 159 161 8

164 161 167 164 161 147 158 158 159 150 153 139 153 163 153 157 8

192 179 174 184 179 178 187 167 178 215 201 212 168 245 157 188 23

112 105 98 114 111 107 108 87 107 150 136 148 98 180 89 117 26

78 74 72 73 65 62 72 68 64 72 63 72 60 64 63 68 5

825 840 855 789 805 864 841 860 855 746 731 754 842 716 857 812 52

197 197 200 195 197 198 192 199 190 200 197 203 191 193 196 196 4

155 150 153 155 147 138 149 158 148 116 127 126 151 129 136 143 13

155 247

206 183

165 144

140 134

257 252

277 179

64 61

759 629

164 188

120 118

0.23 0.24 0.23 0.26 0.47 0.56 0.27 0.21 0.23 0.22 0.23 0.28 0.30 0.23 0.22 0.28 0.10

164 174 184 171 162 157 188 199 197 188 177 191 190 195 205 183 15

269 262 263 274 228 270 252 251 268 251 252 252 232 340 337 267 32

175 175 178 170 162 172 172 166 168 175 171 161 175 196 178 173 8

123 123 127 138 129 129 120 120 125 115 119 117 134 125 121 124 6

225 223 211 224 199 251 202 189 197 172 214 212 221 159 168 204 25

291 282 256 269 277 338 224 220 244 198 319 262 213 182 176 250 48

54 53 58 52 53 60 56 56 57 51 55 53 57 62 51 55 3

700 714 728 712 756 667 759 719 720 710 677 699 755 835 785 729 43

128 132 130 134 142 140 135 133 138 149 135 125 140 137 125 135 7

118 111 121 111 121 121 109 120 127 128 115 117 120 127 119 119 6

0.27 0.45 0.15 0.41 0.08

248 175 317 285 215

226 209 87 109 70

137 199 110 122 96

111 123 108 137 84

155 145 204 194 333

192 116 159 178 229

55 51 50 55 38

661 765 388 674 358

131 154 186 183 158

108 112 117 120 84

Please cite this article as: Waksman, S.Y., et al., Moulded wares production in the Early Turkish/Beylik period in Western Anatolia: A case study from Ephesus and Miletus, Journal of Archaeological Science: Reports (2015), http://dx.doi.org/10.1016/j.jasrep.2015.11.015

S.Y. Waksman et al. / Journal of Archaeological Science: Reports xxx (2015) xxx–xxx

7

Table 1 (continued) SiO2

Al2O3

MgO

MnO

(Na2O)

(P2O5)

Zr

Sr

Ephesus group c/4 (n = 41, Waksman, 2014) m 1.50 11.19 1.020 4.24 σ 0.33 0.47 0.019 0.18

52.72 0.85

23.59 0.57

3.68 0.35

0.1374 0.0162

1.27 0.17

0.29 0.05

228 20

163 10

Ephesus group b/2 (n = 17, Waksman, 2014) m 3.90 8.42 0.973 3.99 σ 0.89 0.21 0.037 0.20

55.82 1.07

21.05 0.79

3.85 0.33

0.0985 0.0053

1.31 0.12

0.25 0.06

177 10

267 24

Lab nb.

Desc.

CaO

Fe2O3

TiO2

K2O

Rb

Zn

Cr

Ni

(La)

Ba

V

Ce

171 8

152 6

185 27

114 27

71 13

810 64

203 8

144 14

182 10

123 7

196 23

242 54

61 14

727 48

141 5

124 9

Abbreviated description of samples: lrt: local reference, tripod stilt; lrb: local reference, biscuit fired ware; lro: other local references; mo: mould; mw: moulded ware; gmw: glazed moulded ware. Major and minor elements are given in oxides weight %, trace elements in parts per million (ppm); m: mean, σ: standard deviation, n: number of samples. Elements between brackets are indicative; data with an asterisk were not taken into account in the calculation of m and σ.

period, even within a category of wares (see especially the case of Pergamum in Waksman, 2014). Conversely, the same clay resources may be exploited over long periods, and used for various categories of objects, as seen here with the example of (table) moulded wares manufactured in a cooking ware fabric. Miletus is no exception, as the reference group defined by Mommsen for the Kalabaktepe workshop in Miletus (Akurgal et al., 2002, 121–122: “Herkunftsgruppe A, Milet – Kalabaktepe-Werkstätten”) can be seen, in spite of the limited number of chemical elements common to Lyon and Bonn laboratories, to be very different from ours (due e.g. to its much lower Cr and Ni contents). On the other hand, another group from Bonn tentatively attributed to Miletus (Akurgal et al., 2002, 126: “Herkunftsgruppe D, Milet(?)”) is not so different, but it is difficult to go beyond this point due to its heterogeneity and in the absence of intercalibration procedure (see e.g. Waksman, 2006).

4.2.1. Slips According to SEM images, all the samples have a slip under the glaze, except for sample BZY933 whose glaze covers the ceramic body directly. The slips are white, of clayey type, and include a significant proportion of siliceous inclusions (mainly quartz), usually higher than in the bodies (Fig. 6a). They are part of the “clay-rich, coarse-grained slips” category as introduced by Capelli and Cabella (2007). The chemical compositions of all the slips show high silicon contents (c. 65–75 wt.% SiO2), and aluminium contents which confirm their clayey nature (c. 11–22 wt.% Al2O3). It may be possible to distinguish two types of slips, according to the microstructure of their interstitial matrix and to the contents in some elements. In samples of Miletus and Ephesus c/4 groups, the interstitial clayey matrix of the slip, located between the inclusions, shows a sheeted structure (Fig. 6c). The chemical compositions of the slips of Ephesus group c/4 have lower aluminium oxide contents (c. 12 wt.% Al2O3 in average) and also higher contents of lead oxide (c. 11 wt.% PbO, vs. 6 wt.% PbO in Ephesus group b/2) (Table 2). Lead may have diffused in larger amounts from the glaze due to the higher porosity of these slips. The samples of Ephesus group b/2 show a more vitreous matrix, particularly close to the interface with the glaze (Fig. 6d), which may explain a lesser diffusion of the lead from the glaze. On the other hand, it suggests that vitrification occurred during the second firing, by interaction with the fluxing agents of the glaze, rather than during

4.2. Slips and glazes Among the moulded wares, seven glazed samples were analysed under the SEM. Three of them belong to Ephesus group b/2 (BZY373, 931, 956), two to Ephesus local group c/4 (BZY932, 934), one to Miletus local group (BZY997) and the last one (BZY933) is a marginal sample found in Ephesus (Table 2).

Table 2 Semi-quantitative SEM-EDS analysis of the glazes (top) and the slips (bottom), in wt.%. Sample code/group

Colour

Thickness (μm)

CaO

Fe2O3

TiO2

K2O

SiO2

Al2O3

MgO

MnO

Na2O

P2O5

PbO

(SnO2)

CuO

BZY373 (Ephesus b/2) BZY931 (Ephesus b/2) BZY956 (Ephesus b/2) BZY932 (Ephesus c/4) BZY934 (Ephesus c/4) BZY933 (marginal) BZY997 (Miletus local) Ephesus b/2 (m) (n = 6) Ephesus b/2 (σ) Ephesus c/4 (m) (n = 11) Ephesus c/4 (σ)

green green green green yellow dark green green

80 80 40 60 40 300 110

0.48 0.79 1.53 0.95 0.24 0.51 0.35 0.85 0.46 0.75 0.41

0.81 0.81 1.23 0.62 0.52 0.83 0.67 1.48 1.87 0.74 0.39

0.17 0.17 0.20 0.18 0.12 0.10 0.15 0.13 0.08 0.16 0.05

0.20 0.46 0.79 0.21 0.09 0.12 0.35 0.34 0.28 0.24 0.19

32.92 37.58 45.27 30.93 26.75 26.46 27.75 36.54 5.37 33.43 4.96

2.79 4.14 5.08 4.75 4.72 2.68 3.47 3.44 0.99 4.83 0.86

0.59 0.58 0.59 0.58 0.52 0.80 0.70 0.53 0.09 0.93 0.51

0.24 0.11 0.08 0.06 0.13 0.12 0.12 0.36 0.57 0.34 0.73

0.39 0.43 0.58 0.56 0.55 0.60 0.66 0.46 0.06 0.65 0.22

0.24 0.10 0.03 0.15 0.15 0.17 0.10 0.15 0.08 0.28 0.21

57.04 51.86 41.53 58.91 65.94 64.96 63.36 53.78 8.16 56.23 6.07

0.05 n.d. n.d. n.d. n.d. n.d. n.d. 0.01 0.02 0.04 0.06

4.09 2.97 3.12 2.10 0.28 2.65 2.31 1.93 1.65 1.38 1.27

Semi-quantitative SEM-EDS analysis of the slips (wt.%). Sample code/group

Thickness (μm)

CaO

Fe2O3

TiO2

K2O

SiO2

Al2O3

MgO

MnO

Na2O

P2O5

PbO

CuO

BZY373 (Ephesus b/2) BZY931 (Ephesus b/2) BZY956 (Ephesus b/2) BZY932 (Ephesus c/4) BZY934 (Ephesus c/4) BZY997 (Miletus local) Ephesus b/2 (m) (n = 6) Ephesus b/2 (σ) Ephesus c/4 (m) (n = 11) Ephesus c/4 (σ)

130 230 90 170 85 110

0.38 0.60 1.53 1.17 0.61 1.14 0.61 0.48 0.59 0.31

1.80 1.52 3.26 1.37 0.75 1.82 2.19 0.97 0.85 0.23

0.87 0.82 0.77 0.84 0.47 0.94 0.77 0.10 0.52 0.12

3.34 2.85 2.75 3.11 2.14 4.34 2.96 0.42 2.32 0.45

66.84 69.38 64.39 68.78 75.07 64.66 65.90 2.93 69.49 4.53

22.27 19.96 19.87 13.32 11.05 18.34 19.07 2.41 12.59 1.89

0.95 0.93 1.93 0.61 0.50 1.31 1.10 0.41 0.72 0.16

0.14 0.07 0.06 0.02 0.08 0.08 0.10 0.03 0.11 0.04

0.58 0.61 0.77 0.99 0.29 0.74 0.72 0.12 0.74 0.19

0.10 0.09 0.11 0.11 0.16 0.02 0.10 0.04 0.26 0.15

2.14 2.83 4.15 9.38 8.63 6.25 6.07 3.77 11.46 2.43

0.59 0.33 0.39 0.29 0.27 0.36 0.40 0.10 0.35 0.06

n.d.: below detection limits; n: number of samples, including both moulded and sgraffito wares; m: mean; σ: standard deviation.

Please cite this article as: Waksman, S.Y., et al., Moulded wares production in the Early Turkish/Beylik period in Western Anatolia: A case study from Ephesus and Miletus, Journal of Archaeological Science: Reports (2015), http://dx.doi.org/10.1016/j.jasrep.2015.11.015

8

S.Y. Waksman et al. / Journal of Archaeological Science: Reports xxx (2015) xxx–xxx

Fig. 6. BSE images of cross sections of glazed wares representative of the types discussed in the text: (a) Ephesus c/4 sample BZY934, (b) marginal sample BZY933 showing feldspars enriched with lead, (c) Miletus local sample BZY997, (d) Ephesus b/2 sample BZY956 (gl: glaze; ws: white slip; cb: ceramic body); binary diagram showing SiO2 contents in the glaze (without lead and colouring oxides) vs. SiO2 contents in the slip (without lead).

the first firing. It is difficult to know at this stage of the study why vitrification started in these slips and not in those of Ephesus group c/4, which are comparatively poorer in aluminium. It may be due to higher firing temperatures and/or to differences in the mineralogy of the interstitial clays. 4.2.2. Glazes The glazes are all homogenous without significant inclusion and only the glaze of sample BZY933 contains some bubbles. They are also

similar in their chemical compositions, very rich in lead oxide (≥ 40% PbO) and poor in alkali contents (≤ 1 wt.% NaO + K2O), lead being thus the main fluxing agent (Table 2). According to Tite et al. (1998), these glazes are classified as “high lead, low alkali”. The two other main components of the glazes are silica (SiO2 ≥ 25 wt.%), and alumina which acts as a stabilizing agent introduced in the glazing mixture either by the addition of clays or by the superficial dissolution of the ceramic body or slip on which the glaze is applied (Kiefer, 1985; Tite

Please cite this article as: Waksman, S.Y., et al., Moulded wares production in the Early Turkish/Beylik period in Western Anatolia: A case study from Ephesus and Miletus, Journal of Archaeological Science: Reports (2015), http://dx.doi.org/10.1016/j.jasrep.2015.11.015

S.Y. Waksman et al. / Journal of Archaeological Science: Reports xxx (2015) xxx–xxx

9

et al., 1998). In these high lead glazes, copper metallic oxides are responsible for the green colour with CuO contents above 2 wt.% (Molera et al., 1997; Paynter et al., 2004; Tite, 2011; Kırmızı et al., 2015). Due to the transparency of the glazes, the green colour may vary according to the colour of the underlying surface: light green when the glaze is applied on a white or cream coloured slip, dark green when applied on a grey coloured body, as it is the case for sample BZY933. Sample BZY934 has a yellow coloured glaze in spite of a low iron oxide content (0.52 wt.% Fe2O3), very high lead contents (c. 65 wt.% PbO) being also possibly responsible for the yellowing of the glaze (Picon et al., 1995). Our results match previous SEM analyses carried out on glazes produced in Ephesus at the same period (Okyar, 2010; Okyar et al., 2011). Only samples BZY934 (Fig. 6a), and especially BZY933, show crystals at the slip/glaze or body/glaze interface (Fig. 6b). These crystals have an euhedral morphology whose elemental compositions may correspond to feldspars enriched with lead, formed in high lead glazes during the firing process (Molera et al., 1993; Tite et al., 1998; Molera et al., 2001; Waksman et al., 2008). Their abundance in sample BZY933 suggests a single firing, unlike the other samples whose glaze was applied on a biscuit-fired slip. Attempts to identify the initial constitution of the glazing mixture were made using the method proposed by Hurst and Freestone (1996) (see also Tite et al., 1998; Walton, 2004; Waksman et al., 2008; Walton and Tite, 2010). On the one hand, this mixture can be created by a lead compound, introduced as galena (PbS) or lead shot, which interacts directly with the ceramic body or slip components to form the glaze during firing. On the other hand, the production of glaze can be realized with a mixture of a lead compound and of silica in the form of quartz sand, ground quartz or chert pebbles. In both cases, clay and organic materials, such as gum or starch, may be added in order to maintain the lead and silica particles in suspension, to give plasticity to the suspension and to improve the binding between the glazing mixture and the underlying slip or body (Parmelee, 1973; Picon et al., 1995; Tite et al., 1998; Waksman et al., 2008). The proposed method allows to discriminate both types of production by subtracting the lead oxides and colouring oxides contents in the glazes compositions. If the resulting composition is close to the one of the slip or body, then it is conceivable that a lead compound was applied directly or was mixed with the same clays used for the slip or body. If the silica content is higher in the glaze than in the slip or the body (recalculated without lead), it seems more probable that the glaze was produced with a silica– lead mixture, and possibly with other ingredients. Fig. 6e shows the comparison between silica contents in the slip (or in the body for sample BZY933) and in the glaze, recalculated after removal of the contribution of lead and colouring oxides. It also includes some other lead-glazed samples of different archaeological types (especially polychrome sgraffito wares), which belong to the same productions (Ephesus groups c/2 and b/4: Sauer and Waksman, 2005; Waksman, 2015) and to the same technical tradition. It seems clear that all our samples, except those of Ephesus group c/4, were covered by a lead–silica glazing mixture. Concerning Ephesus group c/4, the use of this method seems less obvious because the SiO2 contents are not systematically higher in the glaze compositions. However, it is not possible to affirm with certitude that the glaze was produced with a lead compound only, or with a lead–slip mixture, since the glazes compositions after subtraction of the contributions of lead and colouring oxides are relatively different from those of the slips (also recalculated after subtraction of the lead content).

group “Ephesus group b/2”) diffused its products to the two above mentioned sites. The chemical data suggest that this group, possibly located in the surroundings of Ephesus, was significant regionally, as most moulded wares found in Miletus belong to it, and not to the local one, in spite of the fairly large number of moulds found in Miletus itself. It seems thus clear that these wares circulated, and could also appear as imports in sites which manufactured them. On the other hand, we have no evidence so far of the import of moulds, which are local to each production site in the present study. Similarities in the designs of moulded decorations however suggest that craftsmen, moulds, stamps or models circulated. Other evidences, such as a mould found in Sardis, or the large number of moulded wares found in a hamam in Manisa (Gök Gürhan, 2011), suggest that several workshops in the region manufactured these wares and call for further research. Further research will also replace moulded wares in the larger picture of pottery production and dining habits at the early Turkish period in Western Anatolia. It may rely on the definition of chemical reference groups, based on the analysis of local reference material such as pottery wasters and kiln furniture. Previous Ephesian reference groups were reconsidered, one of them (“Ephesus group b/2”) being now considered “regional” rather than local stricto sensu (as is the case of “Ephesus group c/4”). A new reference group was defined for Miletus, which differs from the one defined for wares of the classical period (Akurgal et al., 2002), thus calling again for caution in the diachronic use of chemical reference groups (Waksman, 2014). Moulded wares are among the new types of pottery emerging in several pottery workshops in Western Turkey in the wake of the Turkish conquest. They present new techniques of manufacture when compared to the previous Byzantine productions, or re-introduce earlier ones as is the case for moulds. In terms of slips and glazes, the moulded wares considered here show the association of clayey slips, and of highlead glazes which perpetuate the late Byzantine tradition of glaze “recipe” (Armstrong et al., 1997; Waksman, 2005; Kırmızı, 2012; Kırmızı et al., 2015). For all samples except a marginal one, the lack of crystals at the glaze/slip interface suggests a double firing, which is consistent with the presence of biscuit-fired wasters in the archaeological contexts. The slips show more or less vitrified structures depending on the production considered. However, it is not yet possible to define whether it is due to differences in the firing process (and especially in firing temperatures) and/or in the mineralogy of the clays used for the slips. The evolution towards synthetic slips and opacified glazes will occur slightly later, as seen in the ongoing study of Miletus ware productions (Burlot et al., forthcoming). The introduction of moulded wares may thus be seen as one of the first stages of the transition from Byzantine to Turkish pottery production, whose more advanced stages correspond to the best known Iznik and Kütahya wares (Aslanapa, 1969; Atasoy and Raby, 1989; Henderson, 1989; Tite, 1989; Okyar, 2002; Paynter et al., 2004). The gradual diffusion of pottery manufacturing techniques associated to the Islamic world would certainly deserve further investigations, as moulded wares are already found in Eastern Turkey, Syria and Irak in the 8th–9th c. AD (Rousset, 2001; Orssaud, 2001; Mulder, 2014). But such study would only be speculative unless we found it on well-defined archaeological contexts, safe attributions to workshops, and proper technological characterization. We hope the present paper can contribute to this wider perspective.

5. Concluding remarks and perspectives

This study was funded by the French National Research Agency (ANR) through the POMEDOR project, and we acknowledge the support of the ANR under reference ANR-12-CULT-0008. We would like to thank: the Directors of the archaeological excavations at Ephesus and Miletus; the Turkish Ministry of Culture and the Museums of Miletus and Ephesus for giving us permission to study the samples; B. Aydıl

This study confirms the local production of moulded wares in Ephesus and Miletus at the beginning of the Turkish period. It shows the introduction of this type of ceramics in the local repertoire of several sites in Western Anatolia, one of which (corresponding to chemical

Acknowledgements

Please cite this article as: Waksman, S.Y., et al., Moulded wares production in the Early Turkish/Beylik period in Western Anatolia: A case study from Ephesus and Miletus, Journal of Archaeological Science: Reports (2015), http://dx.doi.org/10.1016/j.jasrep.2015.11.015

10

S.Y. Waksman et al. / Journal of Archaeological Science: Reports xxx (2015) xxx–xxx

and O. Tuncer, S. Demir and the Ephesus team for their contribution to the drawings; the staff of the “Laboratoire de Céramologie” in Lyon. Drawings J. Burlot, S.Y. Waksman and B. Böhlendorf-Arslan, DAO J. Burlot, photos S.Y. Waksman, SEM photos J. Burlot, plates layout S.Y. Waksman.

References Akurgal, M., Kerschner, M., Mommsen, H., Niemeier, W.-D., 2002. Töpferzentren der Ostägäis: archäometrische und archäologische Untersuchungen zur mykenischen, geometrischen und archaischen Keramik aus Fundorten in Westkleinasien. 3. Ergänzungsheft zu den Jahresheften des Österreichisches Archäologisches Institut Wien. Armstrong, P., Hatcher, H., Tite, M.S., 1997. Changes in byzantine glazing technology from ninth to thirteenth centuries. In: d'Archimbaud, D. (Ed.), Acte du VIème Congrès International sur la Céramique Médiévale en Méditerranée, Aix-en-Provence, 13–18 novembre 1995. Narrations, Aix-en-Provence, pp. 225–229. Aslanapa, O., 1969. Pottery and Kilns from the Iznik Excavations. In: Aslanapa, O., Naumann, R. (Eds.), Forschungen zur Kunst Asiens: In Memoriam Kurt Erdmann. Baha Matbaasi, Istanbul, pp. 140–146. Atasoy, N., Raby, J., 1989. Iznik, the pottery of Ottoman Turkey. Alexandria Press, London. Böhlendorf-Arslan, B., 2002. Die Beziehungen zwischen byzantinischer und emiratszeitlicher Keramik. In: Doğan, N.Ş. (Ed.), Ortaçağ Anadolu Semineri, Dr. Aynur Durukan'a Armağan, Ankara, 8–9 Mayis 2000. Hacettepe Üniversitesi, Ankara, pp. 135–156. Böhlendorf-Arslan, B., 2004. Glasierte byzantinische Keramik aus der Türkei. Ege Yayinlari, Istanbul. Böhlendorf-Arslan, B., 2008. Keramikproduktion im byzantinischen und türkischen Milet. Ist. Mitt. 58, 371–407. Budak Ünaler, M., 2013. “Fine-sgraffito”, ”Aegean ware” from Anaia: An Analytical Approach (Ph.D. Thesis) Izmir Institute of Technology, Turkey. Burlot, J., Waksman, S.Y., Bouquillon, A., Vroom, J., Böhlendorf-Arslan, B., Japp, S., 2015. Etude des techniques céramiques des premières productions turques d'Anatolie occidentale forthcoming. Capelli, C., Cabella, R., 2007. The archaeometric study of white slips: a contribution to the characterization of the Medieval Mediterranean productions. In: Waksman, S.Y. (Ed.), Archaeometric and Archaeological Approaches to Ceramics: Papers Presented at EMAC ‘05, 8th European Meeting on Ancient Ceramics, Lyon, 2005. British Archaeological Reports, S1691, London, pp. 155–159. Dalaman, Y., 2015. Ephesus during the Beylik and Ottoman Periods (14th‐–16th c). In: Ladstätter, S (Ed.), Die Türbe im Artemision. Ein frühosmanischer Grabbau in Ayasuluk/Selçuk und sein kulturhistorisches Umfeld. Sonderschriften des Österreichischen Archäologischen Instituts 3. Wien, pp. 9–16. Demirci, Ş., Caner-Saltik, E.N., Böke, H., Bakirer, Ö., Türkmenoğlu, A., Parman, E., 1996. Characteristics of some Medieval glazed ceramics in Anatolia. In: Dermici, Ş., Özer, A.M., Summers, G.D. (Eds.), Archaeometry 94, Proceedings of the 29th International Symposium on Archaeometry, Ankara, 9–14 May 1994. Tübitak, Ankara, pp. 609–615. Doğer, L., 2013. Daskyleion II — Hisartepe/Daskyleion Kazisi Bizans Seramikleri. Ege Yayinlari, Istanbul. Gök Gürhan, S., 2011. Bir Seramik Definesinin Öyküsü. Manisa Belediyesi Kültür Yayani, Manisa. Henderson, J., 1989. Iznik ceramics: a technical examination. In: Atasoy, N., Raby, J. Iznik — the Pottery of Ottoman Turkey 84–87. Alexandria Press, London, pp. 65–69. Hochuli-Gysel, A., 1977. Kleinasiatische glasierte Reliefkeramik (50 v. Chr. bis 50 n. Chr.) und ihre oberitalischen Nachahmungen. Stämpfi & Cie., Bern. Hughes, M.J., Leese, M.N., Smith, R.J., 1988. The analysis of pottery lamps mainly from Western Anatolia, including Ephesus, by neutron activation analysis. In: Bailey, D.M. A Catalogue of Lamps in the British Museum, III. Roman Provincial Lamps. British Museum, London, pp. 461–485. Hurst, D., Freestone, I., 1996. Lead glazing technique from a medieval kiln site at Hanley Swan, Warwickshire. Mediev. Ceram. 20, 13–18. Jones, R.E., 1986. Greek and Cypriot Pottery: A Review of Scientific Studies. British School at Athens, Athens. Kiefer, C., 1985. Caractérisation des tessons types par l'analyse physique et chimique. In: Soustiel, J. La Céramique islamique: le Guide du connaisseur. Office du Livre, Vilo, Paris, pp. 365–378. Kırmızı, B., 2012. Material Characterization of the Late 12th–13th Century Byzantine Ceramics from Kuşadasi Kadikalesi/Anaia (Ph.D. Thesis) Middle East Technical University, Turkey. Kırmızı, B., Göktürk, E.H., Colomban, P., 2015. Colorings agents in the pottery glazes of Western Anatolia: new evidence for the use of Naples Yellow pigment variations during the late Byzantine period. Archaeometry 57 (3), 476–496. Ladstätter, S., 2015. Die Türbe im Artemision. Ein frühosmanischer Grabbau in Ayasuluk/ Selçuk und sein kulturhistorisches UmfeldSonderschriften des Österreichischen Archäologischen Instituts 3. Wien. Mania, U., 2006. Eine neue Werkstatt früher türkischer Keramik — Miletware aus Pergamon. Ist. Mitt 56, 475–501. Mercangöz, Z., 2013. Byzantine Craftsmen — Latin Patrons. Reflections from the Anaian Commercial Production in the Light of the Excavations at Kadikalesi Nearby Kuşadasi. Ege Yayinlari, Istanbul. Molera, J., Pradell, T., Martinez-Manent, S., Vendrell-Saz, M., 1993. The growth of sanidine crystals in the lead of glazes of Hispano-Moresque pottery. Appl. Clay Sci. 7, 483–491.

Molera, J., Vendrell-Saz, M., García-Vallés, M., 1997. Technology and colour development of Hispano-Moresque lead-glazed pottery. Archaeometry 39 (1), 23–39. Molera, J., Pradell, T., Salvadó, N., Vendrell-Saz, M., 2001. Interactions between clay bodies and lead glazes. J. Am. Ceram. Soc. 84 (2), 442–446. Mommsen, H., Japp, S., 2009. Neutroaktivierungsanalyse von 161 keramikproben aus Pergamon und Fundorten der Region. Ist. Mitt. 59, 269–286. Mulder, S., 2014. A Survey and Typology of Islamic Molded Ware (9th–13th centuries) based on the Discovery of a Potter's Workshop at Medieval Bӑlis, Syria. J. Islam. Archaeol. 1 (2), 143–192. Müller-Wiener, W., 1967. Das Theaterkastell von Milet. Ist. Mitt. 17, 279–290. Müller-Wiener, W., Henninger, F.J., Koenigs, W., Kossatz, A.U., 1979. Milet 1977. Vorbericht über die Arbeiten des Jahres 1977. Ist. Mitt. 29, 161–173. Okyar, F., 2002. Characterisation of Iznik ceramics. In: Jerem, E., Biró, K.T. (Eds.), Archaeometry 98, Proceedings of the 31st Symposium, Budapest, Avril 26–May 3 1998, BAR International Series 1043 (II). Archaeopress, Oxford, pp. 621–626. Okyar, F., 2010. A preliminary study of the pottery from a well shaft in Ayasuluk/Ephesos. In: Pfeiffer-Taş, Ş. (Ed.), Funde und Befunde aus des Schachtbrunnen im Hamam III in Ayasuluk/Ephesos. Eine Schamanistische Bestattung des 15. Jahrhunderts., Archäol. Forsch. 16. Verlag der Österreichischen Akademie der Wissenschaften, Wien, pp. 67–72 Okyar, F., Kara, A., Issi, A., Yaygingöl, M., Ünal, D., Doğan, M., Pfeiffer-Taş, Ş., 2011. A study on Medieval pottery from a ceramic Kiln remains in Ayasuluk/Ephesus: compositional and microstructural data. In: Mader, B., Pfeiffer-Taş, Ş. (Eds.), Anzeiger der Philosophisch-Historischen Klasse 146. Verlag der Österreichischen Akademie der Wissenschaften; Wien, pp. 155–178. Orssaud, D., 2001. Les céramiques à glaçure monochrome de Qalat Sem'an (VIIIème–IXème siècles), in. In: Villeneuve, E., Watson, P.M. (Eds.), La Céramique byzantine et protoislamique en Syrie-Jordanie (IVème–VIIIème siècles ap. J.-C.): Actes du Colloque tenu à Amman, 3–5 Décembre 1994. Bibliothèque Archéologique et Historique 159. Beirut, pp. 215–220. Özçatal, M., Yaygıngöl, M., İssi, A., Kara, A., Turan, S., Okyar, F., Pfeiffer Taş, Ş., Nastova, I., Grupče, O., Minčeva-Šukarova, B., 2014. Characterization of lead glazed potteries from Smyrna (Izmir/Turkey) using multiple analytical techniques; part I: glaze and engobe. Ceram. Int. 40 (1), 2143–2151. Parmelee, C.W., 1973. Ceramic Glazes. third ed. Cahners Books, Boston. Parrer, G., 2015. Die archäologisches Befund. In: Ladstätter, S (Ed.), Ein frühosmanischer Grabbau in Ayasuluk/Selçuk und sein kulturhistorisches UmfeldSonderschriften des Österreichischen Archäologischen Instituts 3. Wien, pp. 25–122. Paynter, S., Okyar, F., Wolf, S., Tite, M.S., 2004. The production technology of Iznik pottery — a reassessment. Archaeometry 46, 421–437. Pfeiffer-Taş, Ş, 2010. Funde und Befunde aus dem Schachtbrunnen im Hamam III in Ayasuluk/Ephesos. Eine schamanistische Bestattung des 15 Jahrhunderts, Denkschriften der Philosophisch-Historischen Klasse 389. Verlag der Österreichischen Akademie der Wissenschaften, Wien. Pfeiffer-Taş, Ş., 2011. Reste einer Keramikwerkstatt aus dem 14. Jh. In: Mader, B., PfeifferTaş, Ş. (Eds.), Anzeiger der Philosophisch-Historischen Klasse 146. Verlag der Österreichischen Akademie der Wissenschaften, Wien, pp. 91–154. Picon, M., Lasfargues, J., 1974. Transfert de moules entre les ateliers d'Arezzo et ceux de Lyon. Rev. Archéol. Est 25 (1), 60–69. Picon, M., 1984. Le traitement des données d'analyse. In: Hackens, T., Schvoerer, M. (Eds.), Datation-caractérisation des céramiques anciennes : cours intensif européen, Bordeaux-Talence, 6–18 avril 1981. CNRS, Paris, pp. 379–399. Picon, M., Thiriot, J., Vallauri, L., 1995. Techniques, évolutions et mutations. In: Gaudin, J.C., Luccioni, J., Blistene, B. (Eds.), Le vert et le brun. De Kairouan à Avignon, céramique du Xème au XVème siècle. Musée de Marseille – Réunion des Musées Nationaux, pp. 41–50. Rotroff, S.I., Oliver Jr., A., 2004. The Hellenistic pottery from Sardis : the Finds through 1994. Archaeological Exploration of Sardis Monographs 12. Harvard University Press, Cambridge (Mass.). Rousset, M.-O., 2001. La céramique de Hira à décor moulé, incisé ou appliqué. Techniques de fabrication et aperçu de la diffusion. In: Villeneuve, E., Watson, P.M. (Eds.), La céramique byzantine et proto-islamique en Syrie-Jordanie (IVe-VIIIe siècles apr. J.C.), Amman, 3–5 décembre 1994Bibliothèque Archéologique et Historique 159. Institut Français d'Archéologie du Proche-Orient, Beyrouth, pp. 221–230. Sarre, F., 1935. Die keramik der Islamischen zeit von Milet. In: Wulzinger, K., Wittek, P., Sarre, F. (Eds.), Das Islamische Milet. De Gruyter, Milet III/4, Berlin, pp. 69–88. Sauer, R., Waksman, S.Y., 2005. Laboratory investigations of selected medieval sherds from the Artemision in Ephesus. In: Krinzinger, F. (Ed.), Spätantike und Mittelalterliche Keramik aus Ephesos. Denkschriften der Philsophisch-Historischen Klasse 332Archäol. Forsch. 13. Verlag der Österreichischen Akademie der Wissenschaften, Wien, pp. 51–66. Schneider, G., 2000. Chemical and mineralogical studies of late Hellenistic to Byzantine pottery production in the eastern Mediterranean. In: Zabehlicky-Scheffenegger, S. (Ed.), Rei Cretariae Romanae Fautores Acta 36. Bonn, pp. 525–536 (Bonn). Schneider, G., Japp, S., 2009. Röntgenfluorezenzanalysen von 115 keramikproben aus Pergamon, Ҫandarli, Elaia und Atarneus (Türkei). Ist. Mitt. 59, 287–306. Scott, J.A., Kamilli, D., 1981. Late Byzantine Glazed Pottery from Sardis, in: Actes du XVe Congrès international d’études byzantines, Athènes, 1976. Art et Archéologie vol. 2. Association Internationale des études byzantines, Athens, pp. 679–696. Spieser, J.-M., 1996. Die Byzantinische keramik aus der Stadtgrabung von Pergamon. Gruyter, Pergam. Forsch. 9 Berlin. Tite, M.S., 1989. Iznik pottery: an investigation of the methods of production. Archaeometry 31 (2), 115–132. Tite, M.S., Freestone, I.A., Mason, R.B., Molera, J., Vendrell-Saz, M., Wood, N., 1998. Lead glazes in antiquity: methods of production and reasons for use. Archaeometry 40 (2), 241–260.

Please cite this article as: Waksman, S.Y., et al., Moulded wares production in the Early Turkish/Beylik period in Western Anatolia: A case study from Ephesus and Miletus, Journal of Archaeological Science: Reports (2015), http://dx.doi.org/10.1016/j.jasrep.2015.11.015

S.Y. Waksman et al. / Journal of Archaeological Science: Reports xxx (2015) xxx–xxx Tite, M.S., 2011. The technology of glazed Islamic ceramics using data collected by the late Alexander Kaczmarczyk. Archaeometry 53 (2), 329–339. Vroom, J., 2005. Medieval pottery from the Artemision in Ephesus: imports and locally produced wares. In: Krinzinger, F. (Ed.), Spätantike und Mittelalterliche Keramik aus Ephesos. Denkschriften der Philosophisch-Historischen Klasse 332, Archäol. Forsch. 13. Verlag der Österreichischen Akademie der Wissenschaften, Wien, pp. 17–49 Vroom, J., forthcoming, Medieval Ephesus as a production and consumption centre, In: Madgalino P. and Ladstätter S., (Eds.), Ephesus from Late Antiquity to the Later Middle Ages. Peeters, Leuven. Vroom, J., Fındık, E., 2015. The pottery finds. In: Ladstätter, S. (Ed.), Die Türbe im Artemision. Ein frühosmanischer Grabbau in Ayasuluk/Selçuk und sein kulturhistorisches Umfeld. Sonderschriften des Österreichischen Archäologischen Instituts 3, Wien, pp. 205–292. Walton, M.S., 2004. A Materials Chemistry Investigation of Archaeological Lead Glazes (Ph.D. Thesis) University of Oxford, England. Walton, M.S., Tite, M.S., 2010. Production technology of roman lead-glazed pottery and its continuance into late antiquity. Archaeometry 52 (5), 733–759. Waksman, S.Y., 1995. Les céramiques byzantines des fouilles de Pergame. Caractérisation des productions locales et importées par l’analyse élémentaire par les méthodes PIXE et INAA et par pétrographie (Ph.D. Thesis) University Louis Pasteur, Strasbourg, France. Waksman, S.Y., François, V., 2004–2005. Vers une redéfinition typologique et analytique des céramiques byzantines du type Zeuxippus Ware. Bull. Corresp. Hell. 128-129 (2.1), 87–182. Waksman, S.Y., 2005. A propos de quelques analyses élémentaires de glaçures de céramiques byzantines médiévales. In: Ney, C., Peduto, P., Schvoerer, M. (Eds.), Décor de lustre métallique et céramique glaçurée 7. Centro Universitario Europeo per i Beni Culturali, Scienze e materiali del patrimonio culturale, pp. 83–89. Waksman, S.Y., Wartburg von, M.-L., 2006. Fine sgraffito Ware, Aegean Ware and other Wares: new evidence for a major production of Byzantine ceramics. Report of the Department of Antiquities Cyprus, pp. 369–388. Waksman, S.Y., 2006. The wandering scientist, or the quest for intercalibration. Proceedings of the 34th International Symposium on Archaeometry, 3–7 May 2004. Zaragoza. Institución “Fernando el Católico”, Zaragoza, pp. 563–568.

11

Waksman, S.Y., Bouquillon, A., Cantin, N., Katona, I., 2008. Approche archéométrique des premières « Byzantine glazed white ware » et de productions glaçurées romaines et romaines tardives. In: Biegert, S. (Ed.), Rei Cretariae Romanae Fautorum Acta 40, Durrës, 2004, pp. 531–536. Waksman, S.Y., 2013. The identification and diffusion of Anaia's ceramic products: a preliminary approach using chemical analysis. In: Mercangöz, Z. (Ed.), Byzantine Craftsmen — Latin Patrons. Reflections from the Anaian Commercial Production in the Light of the Excavations at Kadikalesi Nearby Kuşadasi. Ege Yayinlari, Istanbul, pp. 101–111. Waksman, S.Y., 2014. Long-term pottery production and chemical reference groups: examples from Medieval Western Turkey. In: Meyza, H. (Ed.), Late Hellenistic to Mediaeval Fine Wares of the Aegean Coast of Anatolia. Their Production, Imitation and use. Neriton, Warszaw, pp. 107–125. Waksman, S.Y., in collaboration with Carytsiotis, M.-M., 2015. Medieval ceramics from the Türbe: an archaeometric viewpoint. In: Ladstätter, S. (Ed.), Die Türbe im ArtemisionEin frühosmanischer Grabbau in Ayasuluk/Selçuk und sein kulturhistorisches Umfeld. Sonderschriften des Österreichischen Archäologischen Instituts 3, Wien, pp. 293–312. Waksman, S.Y, forthcoming, “Provenance studies”: productions and compositional groups. In: Hunt, A., Oxford Handbook of Archaeological Ceramic Analysis, Oxford University Press; Oxford Wulzinger, K., Wittek, P., Sarre, F., 1935. Das Islamische Milet. De Gruyter, Milet III/4, Berlin. Zabehlicky-Scheffenegger, S., Sauer, R., Schneider, G., 1996. Graue Platten aus Ephesos und vom Magdalensberg. In: Herford, M., Mandel, U., Schädler, U. (Eds.), Hellenistische und Römische Keramik des Östlichen Mittelmeergebietes, Kolloquium Frankfurt, 24.–25. April 1995. Archäologisches Institut der Johann-Wolfgang Goethe Universität, Frankfurt, pp. 41–59. Zachariadou, E.A., 1983. Trade and Crusade. Venetian, Crete and the Emirates of Mendeshe and Aydin (1300–1414). Hellenistic Institute of Byzantine and PostByzantine Studies, Venice.

Please cite this article as: Waksman, S.Y., et al., Moulded wares production in the Early Turkish/Beylik period in Western Anatolia: A case study from Ephesus and Miletus, Journal of Archaeological Science: Reports (2015), http://dx.doi.org/10.1016/j.jasrep.2015.11.015